If you’re not
careful, your data science team could be rapidly snuffed out by a very unlikely
threat. You’ve acted on your vision to bring your organization to a new level
by infusing the best analytic talent, and you’ve secured a promising team of
talented data scientists and analytic managers (and possibly one very talented
consultant); however, how do the others feel about the new members of your
family? In fact, if you’re not careful with the way you structure your
organization, you might have a very unlikely adversary -the existing data
warehouse group. Although your data science team is the new kid on the block,
don’t make the mistake of putting them in competition with their older brother –
the data warehouse group.

Sibling Rivalry

Anyone with two or
more children knows about sibling competition. I have an older brother, so I
know what it’s like growing up in the shadow of someone who came before me. And
even though I’m more charming and intelligent than he is, he came first. So,
I’m always regarded as the little brother, even forty years later. The same
thing happens in organizations.

Unless you’re a
brand new organization, or you’re just forming your enterprise data strategy,
your data science team will be the little brother to the data warehouse
group. With this in mind, you must be sensitive to the dynamic between these
two groups – with specific attention on the incumbents. All too often I see
leaders regale with great pomp and circumstance in honor of their new data
scientists, while inadvertently dismissing the data professionals that run
their organization today. Not only is this not nice, but consider this: the
people you’re shunning have the most power to shut down your new data science
team.

Am I My Brother’s Keeper?

This is difficult to
detect, which is why it’s so important that you understand how this works. Most
likely the data warehouse group has no formal authority over the data science
team and it would be career limiting for the leadership here to make a big
noise. Therefore the strategy, whether intentional or not, will be more
clandestine, and they will probably succeed.

It goes back to
change management 101 and the burning platform. When you roll out a new big
data strategy, one of the biggest challenges you’ll face within the
organization is answering this question, “Why do we need big data
analytics right now?” If you cannot answer this question in a very
compelling way, that has a direct impact on them, you won’t have their
commitment. The resistance will come in the form of complacency: “We’re
fine with what we have now.” And they’re fine because – the existing
business intelligence and data warehouse team takes good care of them.

To affect change,
complacency is a formidable adversary, and if you put the data warehouse group
and the data science team in competition with each other, it’s pretty easy for
the data warehouse group to win this battle. They’re the older brother.
Everybody knows them already. They’re embedded in all the existing workflows.
Everything that comes out of the data science team will be compared to how the
data warehouse group would do it. And the data warehouse group will find ways
to marginalize what the data science team is doing, for the sake of their own
survival.

A Blended Model

The fix here is easy
– just avoid an organizational structure that will put them in competition with
each other. There’s no good reason to create a data science team that sits out
on its own; they should be integrated with the existing data professionals. I
recommend blending the groups so there’s no discernible difference between the
old guard and the new guard. You’ll then have an organization that supports a
unified strategy of traditional and nouveau analytics.

That’s not to say
that reorganization isn’t a good idea. This may be an opportunity to put your
structure in alignment with your size and maturity, especially if you’re facing
a growth crisis. For instance, if your organization is large and weighed down
by process and bureaucracy, it’s time to decentralize. If this is timed with
the birth of a new data science team, don’t make the mistake of keeping the
central data warehouse group where it is and locating the data science team
members in the lines of business. This would not only put them in competition
with each other, but it would still leave you with a monolithic data group to
propagate more process and bureaucracy.

Instead, grow the
existing data warehouse group with data scientists and analytic managers and
spread the entire, unified analytic group into the lines of business while
retaining only a small analytic center of excellence to support your new group.
This eliminates competition between the groups and puts your organizational
structure in alignment with your size and maturity.

Bottom line

In families and
organizations, big brothers and sisters always have the upper hand. Unlike
families however, you have a lot of control over whether or not there will be
sibling rivalry within your organization. If you split the organizations apart,
the incumbent data warehouse group will use its positioning and organizational
leverage to ensure your data science team’s efforts don’t take root. Avoid
internal competition by blending your analytic teams instead of separating
them. If you’re currently planning a party in honor of your new data science
team, order more tables for the data warehouse group – the family’s about to
get bigger.